Agent frameworks & infrastructure
Agent memory & context
4,032 agents · most starred first
Agent memory & context, page 6 of 41
A proposed convention for the .agents/ directory to prevent context bloat and improve agent reasoning in complex codebases.
A trustworthy context layer for AI that remembers, plans and acts — free text into typed commitments, decisions and memories, each with its evidence, in a SQLite file you own. MCP, HTTP, library and typed client.
Local-first, evidence-aware project memory for AI coding agents.
The local-first relational and vector database engine for Cortex-Works. Built with SQLite and LanceDB, it brings Agile methodologies and Scrum ticket management directly to autonomous multi-agent teams.
Local outcome-loop + reward layer for coding agents — remember what worked, ship the next move, measure how it landed. Docs at nimroboai.com
Structured project progress and safe cross-agent handoffs in one local Markdown file
Memory Agent monorepo
🧬 KIP (Knowledge Interaction Protocol) is an open protocol for durable AI agent memory and learning. It connects experience, evidence, knowledge, and skills in a portable, auditable cognitive state, helping agents revise beliefs and improve future decisions through feedback from real outcomes.
Recursive Self-Improvement for AI agents and models, helping your agents self-improve with every run
Context profiler for LLMs and AI agents - used to introspect context contents and reduce token costs
交接 Skill(Jiaojie):跨窗口、跨模型、跨设备、跨语言的 AI 上下文交接工具。换窗口,不失忆;换模型,不重来。Open-source AI context handoff.
A zero-dependency TypeScript memory kernel that gives any agent long-term memory. No training or backpropagation: each output reads a few tag-compatible neighbors, updated by a manual Hebbian rule. Links used often are reinforced; idle ones evaporate by half-life. Based on the paper Local Pheromone Network (arXiv:2606.30669).
Enterprise-grade memory framework for LLMs featuring GPU-optimized inference, vector storage, and automated scaling. Enables hyper-personalized responses through efficient context retrieval and integration.
One infinite session for any agents and harness you have. No more degradation and hallucinations.
Memory that lasts and compounds. MentisDB gives agents durable memory so they do not just remember, they improve over time. It stores append-only thought chains plus a Git-like skills registry, letting skills evolve with experience through versioned, integrity-checked updates and fast retrieval.
Cognitive memory for LLM agents — local-first, and shareable across a team through the cloud.
Lossless storage and search for AI agent sessions, across every agentic client.
A minimal, extensible, agent-native knowledge base that keeps shared context current and inspectable
衔枝 Twig · 叙事记忆引擎 — 面向长程 agent 的开源记忆基础设施:三层叙事记忆 + 情感层(日记/心迹/便签/印章)+ 记忆书前端
Persistent project intelligence and memory plugin for DSH: architecture analysis, cross-session context, TODOs, and optional hybrid retrieval
A linked-list discipline for knowledge, memory and pointer hygiene in large projects: five laws plus portable Python engines (declarative rule table, dry-run health check, link-integrity scan).
Structured, temporal memory for AI agents.
AI memory system combining vector search with temporal knowledge graph. Built-in cognitive engine for agents. Supports memory decay, contradiction detection, and MCP integration.
Shared memory across your team and your AI agents — with judgment about what's worth keeping.
An agent skill to evolve the quality of LLM-Wiki (Graphify) at test time.
Long-running fleet orchestration and memory infrastructure for AI agents. Enables persistent context, shared memory, task execution, and governed coordination across multi-agent systems.
Cite the physical world. An open protocol for verifiable, shared world memory.
An opinionated bootstrap kit for growing a long-lived, personal Hermes agent — memory, character, and a shared workspace.
A comprehensive Model Context Protocol (MCP) server providing AI assistants with powerful task management and agent memories capabilities with project-specific storage.
🧠 High-performance persistent memory system for Model Context Protocol (MCP) powered by libSQL. Features vector search, semantic knowledge storage, and efficient relationship management - perfect for AI agents and knowledge graph applications.
Give your AI agent persistent memory with markdown files. Starter kit from the video: My AI and I Share a Brain.
🧠 A native graph-based memory system for autonomous AI agents, featuring bio-inspired sleep-based memory consolidation and powered by KIP & AndaDB.
An open standard for portable, interoperable AI memory across tools, sessions, and devices.
Persistent Claude Code agents with scheduling, sessions, memory, and Telegram.
Conversations aren't linear — why should AI chats be? The first open-source conversation topology engine. Auto-branching session trees, inherited memory, star-map visualization. Build agents that remember, branch, and grow.
Open-source local-first cognitive memory. AGM-compliant belief revision (49/49 postulates). When a fact changes, downstream beliefs are automatically re-evaluated, not just flagged.
[EMNLP 2026] LightRSI is a modular runtime for lightly deploying recursive self-improvement loops in long-horizon LLM agents.
Persistent memory plugin for Codex. Obsidian-style knowledge base that survives across sessions
Self-hosted continuity memory for persistent AI identities across agents, models, and interfaces.
Code for Paper: AriadneMem: Threading the Maze of Lifelong Memory for LLM Agents
玄霖超脑 · 无量网络 v4 重构版— 66 模块6通道 AI 认知神经系统 · 装一次,你所有的 AI 工具从此共享一个永远不失忆的大脑。跨会话记忆 · 多 Agent 互通 · 自动代码审查 · 零配置
Reliable and Efficient Semantic Prompt Caching with vCache
Git-backed long-term memory for AI coding agents
Local cognitive memory runtime for AI agents — durable recall, temporal versioning, provenance, contradiction governance, and portable encrypted storage. Pure Rust, offline-first, with no required LLM or cloud. Patent pending.
Persistent memory and continuity for long-running AI agents.
Persistent memory, reflection & safety for AI agents. No vector DB, no embedding server — BM25 + multi-hop retrieval runs in-process. MCP server with 10 tools. Apache 2.0.
Structured SOUL evolution framework for AI agents — experience, reflection, governed identity updates, and visual timelines.
A memory layer for AI Agents
Self-hosted AI workspace built around context management. Works solo, scales to teams.
Local-first AI memory — runs offline on any machine with 8 GB+ RAM (SBC, mini PC, laptop, workstation). Zero-loss verbatim archive, knowledge graph, hybrid retrieval. Framework-agnostic, no cloud.
OMem (Office Memory) — local-first memory that automatically ingests your real work (mail, calendar, notes, files) from where it already lives, into a Markdown wiki any AI agent can query. macOS today; Windows coming.
Memory is a long-term memory module for AI agents, providing capabilities for memory extraction, storage, retrieval, and migration for agents running on the openJiuwen framework.
Run an After Action Review after any AI session and write concrete lessons so your agent stops repeating mistakes. A manual Reflexion loop, built on the US Army's AAR method.
Memory and Context Orchestration for Coding Agents
Turn your team's AI coding sessions, GitHub code, and docs into one shared, searchable memory — a self-hosted git truth store you query right from your editor over MCP.
Comparison of memory systems for agents
Local-first memory engine for coding agents (Claude Code/Cursor/Codex): SQLite+FTS5 retrieval, versioned bi-temporal memories, full-session capture, zero dependencies, MCP server
Contextual Memory Virtualisation (CMV): git-like snapshots, branching and trimming for AI coding agent context. Demonstrated with Claude Code.
Agent memory platform.
比Hermes Agent更优的五层记忆法 — AI记忆与智慧升华系统,自动沉淀、跨记忆搜索、周期性提醒、技能自生成
Shared memory + orchestration for your coding agents — one MCP server, persistent vector memory, agent registry
A lightweight persistent-memory companion for OpenCode & Claude
CLI + Local MCP - A shared structured memory store across Claude Code, Cursor, Windsurf, Antigravity, and every MCP client. Semantically queryable.
Open-source CLI and Agent plugin suite for EverMe — cross-device, cross-Agent personal memory for AI Agents.
Plain-Markdown persistent memory for AI coding agents. Long-term facts, daily logs, scratchpad, and semantic search — works with pi, Claude Code, and any LLM agent.
The memory and context-management layer for production AI agents. 95.6% on LongMemEval, 93.2% on LoCoMo, with an open harness. Python + JS SDKs, 24 framework integrations, MCP server.
The open-source experience learning layer for AI agents.
A temporal context graph, a memory API, retrieval primitives, and a multiple-platform integration mesh — designed to be embedded into any host process.
✨🦋 illufly - 【幻蝶】基于记忆蒸馏、资料检索的自我进化智能体
Continuous memory + self-improvement for AI agents. Learns how you work, recalls it automatically, no commands. Claude Code & Codex.
Schema-as-code memory for AI agents in Obsidian: typed cards, entity dedup, link repair, update-in-place, and Ebbinghaus decay. Plain Markdown you own — a Claude Code skill.
Synapse — Temporal knowledge graph memory for AI agents. Self-hosted FalkorDB + Graphiti with hippocampus-layer memory management.
Context management proxy for AI coding agents: verbatim archive, on-demand recall, and saved engine state on supported backends.
Wenlan is a knowledge base for the AI-native age. Your AI agents capture what they learn, Wenlan keeps it current and distills it into source-cited wiki pages you can trust
The Cursor10x MCP is a persistent multi-dimensional memory system for Cursor that enhances AI assistants with conversation context, project history, and code relationships across sessions.
A cognitive memory system for AI agents. Single SQLite file. MCP server included.
Persistent memory for AI coding agents: automatic capture, explainable recall, knowledge consolidation, privacy controls, and portable offline storage. One Python file, zero dependencies.
Long-term, cross-project memory for AI coding agents. Your own Obsidian vault as the source of truth. Daemonless and without opaque databases, your memory belongs to you.
Self-hosted memory server for AI agents. Your brain is a git repo, served over MCP and REST.
Native macOS workspace with one local knowledge graph across Claude Code, Codex, OpenCode, and Cursor.
Durable, local-first, explainable memory for coding agents. A Rust CLI memory substrate built on the franken-stack (Asupersync, FrankenSQLite, Frankensearch, FrankenNetworkX).
WayLog - Save & Export AI Chat History. A local-first extension that turns your fleeting AI conversations into a permanent, git-friendly knowledge base.
Persistent local memory for AI coding agents — Claude Code, Codex CLI, Cursor, any MCP client. Temporal knowledge graph, procedural memory, AST codebase ingest, cross-project analogy. LongMemEval R@5 95.1%, LoCoMo 0.607, BEAM 1M 0.448. 74 MCP tools, 9 IDEs, 100% local.
Cognitive memory engine for AI agents — temporal decay, contradiction detection, autonomous consolidation, knowledge graph, ANN recall via HNSW. Embeddable Rust library with Python bindings; powers yantrikdb-server (HTTP gateway, MCP server, openraft cluster). Apache-2.0.
Synapse — Temporal knowledge graph memory for AI agents. Self-hosted FalkorDB + Graphiti with hippocampus-layer memory management.
Source-grounded MCP memory runtime and Obsidian-compatible vault template for LLM agents.
Agentic event-venue operator demo built on MongoDB Atlas. Showcases three-layer memory (long-term, short-term, shared), hybrid retrieval with Voyage AI multimodal embeddings, optional Langfuse traces, and persona-segmented agent decisions during a rain delay scenario at a.
Pensieve: never lose the thread. A local MCP server and Mac app that links your files, repos, agent sessions, tickets and messages into one thread per task, for you and your AI coding agents.
Give your agent a subconscious: mention it and it comes to mind, with what is connected to it, and the lessons carry over. Local plain-Markdown memory for Claude Code, Codex, Cursor and any MCP app. Every answer cites its page. No API key, no server.
An intelligence layer grounding autonomous agents in verified, real-time knowledge at scale.
A sophisticated biologically inspired memory system for AI agents. Provides organic, high quality, persistent memory with self-maintenance strategies.
A memory layer for AI agents that organizes itself
HippocampAI Autonomous Memory Engine for LLM Agents
Persistent memory, decision-making, and governance for AI agents
Cognee Rust - fast and performant AI memory engine
Context optimization for AI coding agents — lower token cost and latency without sacrificing accuracy.
Multi-agent memory governance demo using Caura + OpenClaw. Three agents (Sales, Legal, Admin) share one memory backend with hard fleet boundaries enforced at the query layer. Shows scoped writes, blocked recall, cross-fleet synthesis, and conflict detection.
Memory, working state, heartbeat, and bounded autonomy for long-running AI agents.
Open-source persistent memory MCP server and self-hosted AI workspace for Claude Code, Codex and OpenCode. SynaBun Assistant coordinates their agents.
Local-first cross-session long-term memory for DeepSeek Harness — 22 narrow tools, hybrid FTS5 + local-vector retrieval, evidence gate, provenance audit, SM-2 review. Zero external services, zero API keys.